MEGA-Bench Task - Super CLEVR Scene Understanding — leaderboard

Metric: Task Score (%). Source: huggingface.co. 17 models tracked.

Top models

#ModelScore
1InternVL3-78B71.4
2InternVL3-38B71.4
3InternVL3-14B71.4
4Gemini 2.5 Pro64.3
5InternVL3-8B64.3
6Ovis2-8B64.3
7Gemini 2.0 Flash (Preview)50
8Gemma 3 27B (IT)42.9
9Gemma 3 12B (IT)42.9
10Llama 4 Scout Base35.7
11Gemma 3 4B (IT)28.6

Interactive version: theaggregate.ai/benchmark?slug=mega-bench-task-super-clevr-scene-understanding · How the rankings work · Data refreshed daily, snapshot 2026-07-22.